NOur's Blog

The Gap: Why Deleting Social Media is Only Step One

Thu, 19 Feb 2026 00:00:00 GMT

Let’s be honest. I didn’t delete social media because I wanted to be a monk. I deleted it because I had “brain rot.”

I was spending 6 to 8 hours a day on my phone. My sleep was trash. My productivity was non-existent. I was constantly procrastinating on deadlines, scrolling through other people’s highlight reels instead of doing my own work. It wasn’t just a bad habit; it was a reflex. I would unlock my phone without even thinking, looking for a notification that wasn’t there.

I felt miserable. I felt stressed. I was constantly comparing my behind-the-scenes to everyone else’s public success.

So I deleted the apps.

The Withdrawal Phase

Here is what nobody tells you: The first week is weird. You feel a phantom limb syndrome. You reach for your phone to kill time, and there is nothing there. You feel bored. You might even feel lonely.

But then, something shifts. After about a week, you stop caring about the news. You stop caring about what people are arguing about on Twitter. You stop caring about the “fake dopamine” of likes and views.

You realise you have a massive gap in your day. All that time you claimed you didn’t have? It was there all along.

The Trap of the “Screen Swap”

Deleting the app is not the cure. It’s just the surgery to remove the tumour. The recovery is up to you.

When I first deleted social media, I didn’t magically start writing novels. I just swapped one screen for another. I started watching YouTube. I started playing games. I watched Manchester United matches (which is stressful in its own way, but at least it’s real).

I realised that intentionlessness was the real enemy. Social media just made it easy to be unintentional.

The Truth About “Missing Out”

People are afraid they will miss out if they delete these apps.

Here is the truth: You will miss out. You will miss out on the stress. You will miss out on the jealousy. You will miss out on the feeling of being inadequate because you aren’t on a yacht in Dubai.

But in exchange? You get to know yourself again. You get to find out what you like when no one is watching.

You might feel lonely at first. But trust me, that solitude is better than the noise.

Delete the app. Mind your own business. See what happens.

How to Actually Fill the Gap

If you delete the apps but don’t fill the gap with something real, you will just find another way to waste time. Here is the process that actually works:

Delete the Apps First: Don’t wait until you “feel ready.” Just do it. Even for 24 hours. Rip the band-aid off.
Face Your Projects: The boredom will hit you. Use that energy to tackle the projects you have been postponing. Use the 5-minute rule (commit to just 5 minutes) to start.
Track Your Time: Use a tool like Toggl Track. When you see exactly where your hours go, you can’t lie to yourself anymore.
Rediscover Your Hobbies: Once the noise quiets down, listen to what you actually want to do. Not what the algorithm wants you to watch. Is it reading? Is it sports? Is it journaling? You have to try things to find what resonates with your actual personality.

The Payoff

You won’t feel “happy” immediately. You might even feel bored. But you will feel relief.

You will stop comparing yourself to others. You will start minding your own business. And when you mind your own business, you actually have the time to build a life you don’t need to escape from.

Reclaim your time. It’s the only thing you actually own.

Introducing FileFlow: Streamlining Your Digital Workflow

Mon, 13 Oct 2025 23:00:37 GMT

In a digital world overflowing with files, documents, and data, staying organized is more challenging than ever. Welcome to the official launch of FileFlow, a powerful Ulauncher extension designed to bring order to the chaos and streamline your digital workflow by leveraging the speed of indexed search.

What is FileFlow?

FileFlow is a smart, intuitive file search extension for Ulauncher that uses plocate/locate to find files almost instantly. It’s built for speed and efficiency, and it’s packed with features to help you find what you’re looking for, wherever it is.

Core Features

FileFlow is more than just a simple file search. Here are some of the features that make it an indispensable tool for any power user:

Blazing-Fast Indexed Search

At its core, FileFlow uses your system’s file index to deliver search results in a fraction of a second. Just type s followed by your search term, and watch the results appear instantly.

Directory-Only Search

Need to find a folder? No problem. Use the dir keyword to filter your search and show only directories. This is perfect for quickly navigating your filesystem.

Hardware Drive Scanning

One of FileFlow’s standout features is its ability to search mounted hardware drives. By using the hw keyword, you can extend your search to include external hard drives, USB sticks, and other removable media.

Advanced Hardware and Directory Search

Combine the hw and dir keywords to perform a highly specific search for directories located only on your mounted hardware. It’s the ultimate tool for finding that one specific folder on that one specific drive.

More Than Just Searching

FileFlow also includes a number of other features to improve your workflow:

“Open With…” Menu: Don’t want to use the default application? Press Alt+Enter on a result to bring up a menu of applications to open the file with.
Open Folder Location: Quickly jump to the folder containing a file from the Alt+Enter menu.
Copy All Paths: Need a list of all the files you found? A quick Ctrl+Enter will copy all the result paths to your clipboard.

Get Started with FileFlow

Ready to take control of your digital life? Installing FileFlow is as simple as adding it to your Ulauncher extensions.

Where to Get It

You can download Ulauncher from its official website. Once you have Ulauncher installed, you can add the FileFlow extension from its official page on the Ulauncher extensions website or directly from the GitHub repository.

Thank you for reading!

Unlock Your Second Brain: Introducing InsightMiner

Sun, 17 Aug 2025 23:00:37 GMT

In an age of information overload, we are constantly consuming articles, books, and documents. We highlight, we take notes, but when it comes time to recall a specific piece of wisdom, it’s often buried under a mountain of digital clutter. What if you could turn your personal library into an intelligent, searchable database that not only finds what you’re looking for but also connects ideas and extracts the most important insights?

Here comes InsightMiner, a powerful command-line tool designed to do just that. It’s a local-first, AI-powered system that transforms your collection of Markdown files into a “second brain,” allowing you to ask complex questions and receive concise, relevant answers drawn directly from your own curated content.

What is InsightMiner?

At its core, InsightMiner is a Retrieval-Augmented Generation (RAG) system that runs entirely on your local machine. Unlike cloud-based solutions, it doesn’t require API keys or send your data to third-party services. Your knowledge base remains private and secure.

It’s built for researchers, students, writers, and anyone who wants to derive more value from their reading material. By processing your documents—book highlights, articles, research papers—it creates a sophisticated vector database that understands the meaning behind the words, not just keywords.

How It Works: From Chaos to Clarity

The magic of InsightMiner lies in its two-step process:

Building the Database (python run.py database create): First, the tool reads all your Markdown files from a designated highlights/ folder. It intelligently breaks the text into semantic chunks, generates vector embeddings (numerical representations of the text’s meaning), and stores them in a local ChromaDB database. This process needs to be done only once to index your entire library.
Querying Your Knowledge (python run.py query "your question"): This is where you interact with your second brain. When you ask a question, InsightMiner performs a semantic search to find the most relevant passages from your documents. It then uses a cross-encoder model to re-rank the results for accuracy, extracts the most potent quotes, and finally, generates a summary of the key insights.

The result is a beautifully formatted, easy-to-digest answer that pulls from multiple sources in your library.

An Example in Action

Let’s say you’ve been reading several books on productivity and want to consolidate the key takeaways. Instead of manually sifting through your notes, you can simply ask InsightMiner:

python run.py query "how to be productive"

Here’s a sample of what you might get back:

🔍 Top 5 Results for: 'how to be productive'

📌 Result 1 (Relevance: 1.00)
📚 Source: 18 Minutes
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
To get the right things done, choosing what to ignore is as important as choosing where to focus.
You can be relatively certain that if you decide when and where you’re going to do those things, you’ll actually, reliably and predictably, get them done.
What can you realistically accomplish that will further your focus for the year and allow you to leave at the end of the day feeling that you’ve been productive and successful?
Set your watch, phone, or computer to ring every hour and start the work that’s listed on your calendar.
At the end of your day, shut off your computer and review how the day went,.
Don’t fight yourself to change your behavior in the midst of the wrong environment; just change the environment.

📌 Result 2 (Relevance: 0.97)
📚 Source: 18 Minutes
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
The rule is simple: When you want to do something, focus. When you don’t want to do something, distract.
Distraction, used intentionally, can be an asset.
In contrast to almost everything else in your life, the more you multitask, the worse you are at it. Practice, in this rare case, works against you.
We don’t actually multitask. We switchtask. And it’s inefficient, unproductive, and sometimes even dangerous. Resist the temptation.
Productivity can be achieved only through imperfection.
Stay alert and adapt to changing situations. Keep your eye on the ball, whichever ball that may be.
If you’re having difficulty starting, though, choose your one thing—the one thing that will make the biggest impact.

📌 Result 3 (Relevance: 0.93)
📚 Source: The Time Chunking Method
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
You need to have a reason for doing something in order to make performing that task worth your time and effort.
The less distracted you are, the more productive you’ll be.
If you have superhuman focus in the morning, but the focus of a bored 3-year-old in the afternoon, test the following schedule: Work for 90 minutes Take a 15-minute break Work for 60 minutes Take a 10-minute break Work for 50 minutes Take a 10-minute break.
If laziness is preventing you from using time chunks, that’s a sign that your reasons are not sufficiently compelling to you.

📌 Result 4 (Relevance: 0.75)
📚 Source: Feel-Good Productivity
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
"thinking I was being taught at medical school to questions of happiness,
fulfilment and productivity. So my final exercise brings us full circle, and
involves learning to think about productivity like a scientist. To experiment
with what brings you meaning. And use those experiments to inform the decisions
you make every hour. ‘Alignment experiments’ can help you test theories about
what might bring you closer to alignment in your daytoday decisionmaking. It’s a
process with three stages. First, identify an area of your life where your
actions feel particularly unfulfilling. The results of your eulogy method,
odyssey plan and wheel of life exercises might have helped with this


📌 Result 5 (Relevance: 0.53)
📚 Source: Feel-Good Productivity
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Adopting an experimental mindset. All scientific experiments have an ‘independent variable’, the one thing you change to see what effect it could have. If you were to change one – just one – independent variable in your life, what would it be? And what effect do you think it would have on your situation?
My life has changed. These days, I know that productivity isn’t about discipline; it’s about doing more of what makes you feel happier, less stressed, more energised.
If you can tap in to what makes you feel most energised and alive, you can get anywhere.

════════════════════════════════════════════════════════════════════════════════
                                  ✨ Insights ✨                                  
════════════════════════════════════════════════════════════════════════════════
✨ Top Insights About 'how to be productive' ✨

1. The less distracted you are, the more productive you’ll be.
2. What can you realistically accomplish that will further your focus for the year and allow you to leave at the end of the day feeling that you’ve been productive and successful?
3. My life has changed. These days, I know that productivity isn’t about discipline; it’s about doing more of what makes you feel happier, less stressed, more energised.
4. Productivity can be achieved only through imperfection.
5. You need to have a reason for doing something in order to make performing that task worth your time and effort.

The tool doesn’t just give you a list of documents; it synthesizes the core ideas into actionable insights. This is the true power of InsightMiner—it transforms passive notes into active, usable knowledge.

Why Go Local?

In a world dominated by cloud services, the decision to keep InsightMiner fully local is a deliberate one. It offers several key advantages:

Privacy: Your data never leaves your machine.
No Costs: You are not dependent on paid APIs.
Offline Access: Your knowledge base is always available.
Customization: You have the freedom to swap out the underlying AI models to fit your specific needs.

If you’re ready to take control of your digital notes and build a truly personal knowledge engine, give InsightMiner a try. It’s a step towards working smarter, not harder, and unlocking the collective wisdom hidden in your own reading.

Forged in Chaos: A Manchester United Comeback

Fri, 18 Apr 2025 11:00:37 GMT

“Take me home Take me home, United Road. To the place, I belong. To Old Trafford, To see United. Take me home, United Road.”

You know that chant. You’ve sung it when the crowd was roaring. You’ve whispered it when you were alone in your bedroom at 2 a.m., staring at the ceiling after another crushing loss. And last night? You screamed it , hoarse, tear-streaked, utterly broken , as Harry Maguire’s header crashed into the net and time ran out on Lyon’s dreams. That match wasn’t football. It was pure, unfiltered emotion.

The First Half: A Promise Unfulfilled

We started like a team with purpose. Bruno dancing between midfielders. Casemiro dictating tempo. Mainoo , just 18 years old , looking like he’d been playing top-level European football his whole life. For 45 minutes, we looked like the club we used to be. Sharp. Brave. Believing. Then came the second half. And oh, God… what a descent. A goal. Then another. Then a third. And then , no. Not again. Not this season. Not now. 2-4 down. I turned off the lights. Sat on the floor. Stared at the wall. I thought: This is it. Another collapse. Another chapter written in disappointment. But then… something strange happened. Somewhere deep inside , buried under years of heartbreak, under tweets full of rage, under the weight of expectation , there was still a tiny spark. One percent faith. And somehow? That was enough.

The Turning Point: Casemiro’s Courage

Let’s talk about Casemiro. Not for his tackles. Not for his passes. But for the penalty. The moment the referee pointed to the spot, I didn’t even breathe. I knew what was coming. Bruno stepped up. Calm. Cold. Perfect. 1-4. Then 2-4. Then 3-4. Each goal felt like a miracle. Each one pulled us back from the edge. But it was Casemiro who made the first step. Who won the ball. Who stood tall when everyone else seemed ready to fold. Quiet. Unseen. Always there. He didn’t celebrate. Didn’t shout. Just nodded. Like he’d expected it all along. That’s leadership.

Kobbie Mainoo: The Boy Who Dreamed It

And then… Mainoo. 18 years old. A kid who should’ve been in U18s. Instead, he’s scoring in the Europa League quarter-final second leg… away from home… after we’re down by two goals… in extra time. That finish? Pure instinct. Pure guts. I swear, I saw Wolves’ winner against City flash through my mind , that same look in his eyes. That same refusal to accept defeat. I braced for penalties. I already pictured the agony. The missed spot-kicks. The silence after the final whistle. I wasn’t ready for what came next.

Harry Maguire: The Captain Who Refused to Quit

Seconds later. A corner. A scramble. A flick. And then , Maguire. Not the villain they painted. Not the meme. Not the punchline. Just… a man. Who refused to let this club die. He rose. He headed. He scored. And in that moment? I didn’t just cry. I remembered. Remembered why I fell in love with this club. Not for trophies alone. But for moments like this. When the world says you’re done… …and you find a way to rise anyway. I screamed so loud my dad knocked on the door. I didn’t answer. I didn’t care. I was watching my team do the impossible.

This Wasn’t Just a Win. It Was a Rebirth

Three games left. A semi-final. A final. Opponents tougher than Lyon. Pressure heavier than ever. But now? Now I believe. Because belief isn’t built on wins. It’s forged in chaos. Last year, Amad Diallo’s wonder-goal against City lit a fire. We went on to win the FA Cup. This? This feels bigger. It’s not just about silverware. It’s about identity. It’s about proving , to ourselves, to our rivals, to every doubter who said “they’re finished” , that this club still has soul. Ruben Amorim said it: If you reach the semis, you have to win it. Well… we’re here. And now? We’re not just hoping. We’re hunting.

Bruno Fernandes: The Heartbeat of This Team

Let’s take a moment. For Bruno. Because if you don’t see him as the captain of this team right now , you’re not watching closely enough. He took responsibility when no one else would. He scored the penalty that started it all. He chased every ball. He barked orders. He led. He carried. He’s been the glue holding this fragile squad together all season. I made that pixel art portrait of him because I couldn’t put into words how much he means to us. So I’ll say it now: Thank you, Bruno. For never quitting. For carrying us. For being brave when it hurt most. You don’t need a trophy to be a legend. You just need moments like this. And you’ve given us more than your fair share.

Final Whistle: What Remains

They’ll write headlines. They’ll debate tactics. They’ll dissect errors. But none of that matters tonight. Tonight, we remember:

A boy who scored like a veteran.
A captain who fought like a warrior.
A penalty kick that sparked hope.
A header that brought tears.
A team that refused to die.

Man United didn’t win because they were perfect. They won because they were alive. And sometimes… that’s all you need. To the players: Thank you. To the fans: Keep singing. To Old Trafford: We’re still here. And we’re not letting go.

Bloody hell. What a night. What a club. Onward. 🟡🔴

Pi Day Symphony: A Composition from the Digits of Pi

Fri, 14 Mar 2025 11:00:37 GMT

There’s a quiet magic in the way mathematics whispers through the world, sometimes in spirals, sometimes in stars… and sometimes, in sound.

This Pi Day, I didn’t bake a pie.

I composed one.

Using nothing but the first 1,000 digits of π , that endless, irrational, beautifully unpredictable number , I turned math into melody. The result? A living, breathing piece of music, born from code, shaped by numbers, and meant to be felt.

You can listen here:
🎧 Listen to “Pi as Music” on X (Twitter)

Why Pi? Why Music?

Math and music have danced together since Pythagoras plucked strings and noticed harmonies in ratios. Euler’s formula sings in waves. Fibonacci spirals echo in sonatas.

But pi?
Pi is different.

It doesn’t repeat. It doesn’t settle. It refuses to be tamed.

And yet… it’s everywhere.

In circles. In waves. In the rhythm of your heartbeat.

So I wondered: What if we let pi speak, not as a decimal, but as a song?

Translating Infinity Into Sound

I used Sonic Pi , a beautiful, code-based music tool that turns programming into poetry , to map each digit of π to musical elements. Not arbitrarily. Not randomly. With intention.

Here’s how the numbers became notes:

🎼 Melody

Each digit (0–9) mapped to a note within a changing scale , mostly C major, but shifting subtly to avoid repetition.

0 = C, 1 = D, 2 = E, …, 9 = B

The melody doesn’t loop. It never repeats. Just like pi itself.

⏱️ Rhythm

Digits controlled note lengths:

0 → 0.25s, 1 → 0.5s, 2 → 0.75s, up to 9 → 2s

Longer digits = longer, lingering tones. Short ones = staccato pulses. The rhythm breathes with pi’s chaos.

🎹 Harmony

Every three digits were averaged to pick a chord:

[3, 1, 4] → average = 2.67 → mapped to C major
[5, 9, 2] → average = 5.33 → mapped to F minor

Only major and minor chords were used , clean, emotionally resonant. No dissonance for dissonance’s sake. Just harmony born from structure.

🎛️ Timbre & Texture

Different instruments brought out different layers:

Piano carried the melody , clear, intimate.
Ambient pads swelled with harmonic averages , like mist over water.
Deep bass synth pulsed every fourth digit , grounding the piece in a steady, hypnotic groove.

🔊 Dynamics

Volume followed the digit directly:

0 = softest (pianissimo)
9 = loudest (fortissimo)

A high digit isn’t just a note , it’s a sigh. A cry. A whisper.

The Code Behind the Symphony

Four synchronized threads ran in Sonic Pi:

live_loop :melody do
  pi_digits.each_with_index do |digit, i|
    play scale(:c4, :major)[digit], release: 0.3, pan: rand(-0.5..0.5)
    sleep get_duration(digit) # Maps digit to time value
  end
end

live_loop :bass do
  pi_digits.each_slice(4) do |group|
    note = scale(:c3, :pentatonic)[group[0]]
    play note, attack: 0.1, release: 1.5, amp: 0.8
    sleep 1
  end
end

live_loop :harmony do
  pi_digits.each_slice(3) do |triplet|
    avg = (triplet.sum.to_f / 3).round
    chord = avg.even? ? :c_major : :f_minor
    play_chord chord, release: 4, amp: 0.3
    sleep 4
  end
end

live_loop :percussion do
  digit = pi_digits.tick
  case digit
  when 0..1 then sample :drum_heavy_kick
  when 2..3 then sample :drum_snare_hard
  when 4..5 then sample :drum_cymbal_closed
  end
  sleep 0.25
end

The Price of Personal Choice: An Ethical Dilemma

Mon, 05 Feb 2024 19:00:37 GMT

We all make choices, some small, some life-altering.
That extra slice of cake. The late-night cigarette. The soda instead of water. The skip in exercise. The silence when we should have spoken up.

Most of us live with the quiet consequences. But what happens when those choices lead to something irreversible? When a liver fails. When a heart gives out. When a lung can no longer breathe?

And then… there’s an organ waiting. One that could save a life.

But should that life be yours, if you helped break it?

This is the haunting question at the center of Glenda Cooper’s 1997 essay, “Should those who abuse their bodies pay the price?” It’s not just about medicine. It’s about justice. About fairness. About what we owe each other, and what we owe ourselves.

The Weight of Responsibility

Cooper doesn’t ask whether people should take care of themselves. We all know they should.

She asks something far harder:

If someone knowingly harms their body through persistent, preventable behavior, should they be denied care when they finally need it most?

Imagine two patients needing a liver transplant:

Patient A: A lifelong non-smoker, active, balanced diet, never drank excessively.
Patient B: Diagnosed with cirrhosis after 20 years of heavy drinking, despite multiple warnings from doctors.

Both are dying. Only one liver is available.

Who gets it?

Our instinct might be to choose Patient A. They didn’t cause their own suffering. They’ve been responsible. They deserve it.

But here’s the twist:
What if Patient B has tried to quit? What if they’re now sober? What if their regret is heavier than any guilt we could assign them?

Is fairness about who deserves it most… or about who needs it most?

Ethics vs. Morality: A Crucial Divide

Cooper makes a thoughtful distinction between morality and ethics, one I wish more of us remembered.

Morality is society’s rulebook: You shouldn’t smoke. You should exercise. You must take care of your body.
Ethics is the deeper question: How do we treat people who fall short of the rules?

The danger comes when we confuse the two.

It’s easy to judge.
It’s harder to hold compassion alongside accountability.

When we say, “They brought this on themselves,” we risk turning healthcare into a meritocracy, where worthiness replaces need. Where dignity becomes conditional.

And yet…
Healthcare resources are finite. A single transplant costs hundreds of thousands. There are waiting lists. Families crying in corridors. Doctors making impossible calls every day.

So where do we draw the line?

The Flawed Analogies That Undermine the Argument

Cooper tries to strengthen her case with comparisons, but some land awkwardly.

She compares children eating candy to adults choosing alcohol. But children don’t have full autonomy. Adults do. That’s not negligence, it’s agency.

Then she draws a parallel between doctors allocating organs and Nazis selecting victims.

This is deeply problematic.

One is about saving lives under scarcity.
The other is about exterminating them.

Equating the two doesn’t sharpen the argument, it shatters its moral credibility.

We don’t need shock value to make this point. The truth is painful enough without sensationalism.

The Real Question Isn’t About Blame, It’s About Hope

Perhaps the most uncomfortable truth is this:

We already make these decisions.

Hospitals screen for sobriety before listing for transplants.
Insurance companies deny coverage for complications tied to smoking.
Waitlists prioritize those who show commitment to change.

We aren’t denying care because someone was “bad.”
We’re prioritizing care for those who still have a chance to heal.

It’s not punishment. It’s pragmatism.

But here’s what we must never forget:

People don’t become addicted because they’re weak.
They become addicted because pain, trauma, isolation, or despair led them to find relief in something destructive.

And recovery? It’s rarely linear.

So when we ask, “Do they deserve it?”
Maybe we should also ask:

Can they still become worthy of hope?

A Better Path Forward

I don’t believe anyone should be denied care simply because they made mistakes.

But I also believe we owe it to our systems, and to each other, to design care that encourages responsibility, not shame.

What if, instead of denying treatment, we offered it with conditions?

A year of sobriety before transplant eligibility.
Counseling paired with medical intervention.
Community support, not judgment, as part of the healing process.

This isn’t about withholding compassion.
It’s about extending it in a way that works.

Because medicine isn’t just about fixing bodies.
It’s about restoring dignity.

Final Thoughts: Compassion Without Cynicism

We live in a world that loves to punish failure.
But healing? Healing requires grace.

Let’s not pretend that personal choice exists in a vacuum.
Poverty, mental health, addiction, trauma, they shape our decisions more than we admit.

Yet let’s also not ignore the reality of limited resources.
We cannot pretend we have infinite time, money, or organs.

So what’s the answer?

Not “punish” or “reward.”
Not “they deserve it” or “they don’t.”

But:

How can we help people reclaim their health, without abandoning our duty to use what we have wisely?

Maybe the most ethical thing we can do isn’t to decide who lives or dies.
It’s to create a system where fewer people ever reach that breaking point.

Until then, we sit in the gray.
With trembling hands.
And open hearts.

Thank you for sitting here with me in it.

🫶

Predicting Blood Sugar: A Machine Learning Approach

Fri, 15 Sep 2023 18:00:37 GMT

A Personal Journey Into the Numbers Behind Blood Sugar

Let’s be honest, managing diabetes isn’t just about checking glucose levels or counting carbs. It’s a quiet, daily dance with your body. One minute you’re fine; the next, your sugar spikes, or crashes, without warning.

I’ve seen how overwhelming this can be. That’s why I built the Diabetes Tracker Project: not just as a technical exercise, but as an attempt to bring some calm, clarity, and predictability into a world full of uncertainty.

This isn’t science fiction. It’s real data, over 14,000 blood glucose readings collected every five minutes across five days, and it taught me something powerful: machine learning doesn’t replace human intuition… it amplifies it.

How We Got Here: From Raw Data to Insight

The project began with a public dataset from Jonathan Lifferth’s GitHub repository, which offered continuous glucose monitoring (CGM) data, like what you’d get from a modern wearable sensor.

But raw data? Messy. Overwhelming. Full of timestamps, extra columns, and gaps.

So we cleaned it up:

Removed useless columns like "Unnamed: 0" (yes, they exist more than you’d think).
Grouped readings into 30-minute windows to reflect real-world decision-making.
Created new features: “Glucose Minus 30,” “Glucose Minus 60,” even “Glucose Minus 90”, essentially asking the model, “What did your body do in the last hour?”
Dropped rows with missing values. No guesswork allowed.

We weren’t trying to overcomplicate things. We wanted to listen, to let the data speak in its own rhythm.

The Models We Tested: Who Played Well Together?

We trained five classic regression models, each with its own personality:

Model	Style
Linear Regression	The straightforward one, assumes everything follows a straight line.
Elastic Net	A thoughtful hybrid, balances simplicity and strength.
Random Forest	The team player, makes decisions by listening to many voices.
Gradient Boosting	The perfectionist, learns slowly, correcting mistakes step by step.
Support Vector Regression	The quiet genius, powerful, but sometimes too slow for real life.

And then came the twist: instead of training all models the same way, we tested three different approaches:

First 70% training , Simple chronological split.
Random 70% training , Shuffled randomly (common in ML, but unrealistic for time-series).
Drop 7th reading training , We skipped every 7th reading to preserve temporal structure while still having enough data. This felt closest to real-world usage.

Why? Because diabetes doesn’t care about random sampling. Your sugar doesn’t jump around randomly, it trends, lags, reacts. So our evaluation had to respect that.

The Surprising Winner: Teamwork Makes the Dream Work

Here’s what we found, the most telling numbers from our tests:

Model	MSE - First 70%	MSE - Random 70%	MSE - Drop 7th Reading
lin_model	21.73	21.35	21.46
en_model	21.73	21.35	20.36
rand_model	40.24	17.15	39.93
gb_model	40.26	19.83	39.96
svr_model	90.61	73.42	90.64
(rand + gb)_model	27.87	18.09	16.76 ← Best!

👉 The star performer? A combination of Random Forest and Gradient Boosting, trained using the “Drop 7th reading” method.

With an MSE of just 16.76, this ensemble didn’t just outperform single models, it shattered them in realistic conditions.

What does that mean?

It means we could predict your glucose level 30 minutes ahead, with meaningful accuracy, not by guessing, but by learning from patterns your body has shown before.

Why This Matters Beyond the Numbers

Let me say this plainly:
This isn’t about building a better algorithm. It’s about building a better life.

For someone living with diabetes, knowing their sugar might drop after lunch, or rise after stress, isn’t just helpful. It’s life-changing.

Imagine:

Getting a gentle alert before a nighttime crash.
Adjusting insulin doses based on trend, not just snapshots.
Feeling less anxious because your phone isn’t just showing a number, it’s offering insight.

That’s the promise here.

Machine learning doesn’t need to be complex to be powerful. Sometimes, it just needs to be thoughtful.

What’s Next?

You can explore the full code, datasets, and notebooks on GitHub:
🔗 Diabetes Tracker Project on GitHub

I’m already thinking about next steps:

Adding meal logs or activity data.
Building a mobile interface with simple visual alerts.
Testing with real users, not just datasets.

Because technology should serve people. Not the other way around.

Final Thought

Mathematics, code, and data aren’t cold tools.
They’re bridges, between uncertainty and peace, between chaos and control.

If this project helps even one person feel a little more confident managing their health… then it was worth every late night spent debugging and tuning.

Thank you for reading.
And if you’re living with diabetes, or caring for someone who is, you’re already doing something extraordinary.

Keep going.
You’re not alone.

💛

Visualizing Pi: A Symphony of Digits

Thu, 14 Sep 2023 11:00:37 GMT

Hello, fellow seekers of beauty in the unexpected!

I’ve always believed that math isn’t just about numbers, it’s about stories. And few stories are as timeless, mysterious, and quietly poetic as that of π.

Yes, I know, Pi Day is on March 14th. But honestly? I couldn’t wait. So last week, I let my curiosity take the wheel and turned the first 10,000 digits of pi into something you can see, not just calculate. What emerged wasn’t just a chart or a graph, it felt like a symphony written in circles, color, and quiet rhythm.

A Little Love Letter to Pi

Let’s pause for a moment, not to crunch numbers, but to marvel.

Pi (π) is the ratio of a circle’s circumference to its diameter. Simple enough, right? Roughly 3.14. But here’s the magic: go further. Keep going. Past 3.14159… past 3.1415926535… and you’ll never hit repetition. Never. It goes on forever, irregular, unpredictable, a decimal ocean with no shores.

No fraction captures it. No pattern tames it. And yet, we keep chasing it. Trillions of digits computed. Not because we need them… but because we’re curious. Because there’s something deeply human in wanting to understand the infinite.

And then there’s Pi Day, March 14th. A delightful excuse to eat pie, share fun facts, and celebrate the quiet wonder hiding in plain sight. Who knew a number could bring people together like this?

My Creative Experiment: Turning Digits Into Dance

So I asked myself: What if we gave each digit of pi a voice, and a face?

I wrote a simple Python script (yes, even code can be poetic), fed it the first 10,000 digits, and let it paint.

Each digit became a circle.
Its color? Determined by its value, from soft blues (0) to deep purples (9).
Its size? Gently pulsing, growing slightly whenever the digit changed from the one before it.
Its place? Laid out in neat rows, like pages of an ancient book, left to right, top to bottom.

Here’s the heart of it:

import matplotlib.pyplot as plt

# Load the first 10,000 digits of pi from a file
with open('pi_digits.txt', 'r') as f:
    pi_digits = f.read().strip()[:10000]

# Initialize lists for positions, sizes, and colors
x, y, size, color = [], [], [], []

# Set up the plot
fig, ax = plt.subplots(figsize=(12, 12))

# Loop through each digit and generate visual properties
for i, digit in enumerate(pi_digits):
    if i == 0:
        x.append(0)
        y.append(0)
        size.append(50)
        color.append(int(digit) / 9)
    else:
        if digit != pi_digits[i - 1]:
            size[-1] = 50 + 5 * int(digit)
        x.append(i % 100)
        y.append(i // 100)
        size.append(50)
        color.append(int(digit) / 9)

# Plot the visualization
ax.scatter(x, y, s=size, c=color, cmap='gnuplot')
ax.axis('off')
ax.set_aspect('equal')
plt.show()

The Art of Japanese: Where Language Paints Meaning

Tue, 22 Aug 2023 23:00:37 GMT

Konnichiwa, friends.

Today, I want to share something deeply personal, not just about learning a language, but about falling in love with a way of thinking.

I’ve studied three other languages before Japanese. French for its romance, Spanish for its rhythm, German for its precision. But none of them… none of them made me pause mid-sentence and whisper, “Wait, how did they think this?”

Japanese didn’t just teach me new words. It taught me to see differently.

Writing as Art, Not Just Communication

Let’s start with the characters.

Hiragana flows like ink on silk, soft, curved, alive. Katakana is sharp, deliberate, almost architectural. And kanji? Kanji feels like holding a piece of history in your hand.

Take the katakana フ (fu). Look closely. Doesn’t it look like an open mouth? A breath forming a sound? That’s not coincidence, it’s design. Every stroke was chosen not just for function, but for meaning. In most languages, letters are abstract symbols. In Japanese, they’re tiny paintings.

And when you learn to write them? You don’t just memorize shapes. You trace centuries of thought.

I used to think calligraphy was just beautiful handwriting. Now I understand: it’s meditation in motion.

Kanji: The Logic of Layers

Here’s where my engineering mind fell head over heels.

In Japanese, words aren’t built from random letters, they’re constructed, like circuits or Lego bricks.

Take the word for “train station”: 駅 (eki).

馬 (uma) = horse
又 (mata) = again, or “to cross”
二 (ni) = two

Wait… horses? Two? What?

Actually, 駅 originally meant a place where riders would change horses along ancient routes. Over time, it evolved into “station”, but the meaning still lingers beneath the surface.

This is what makes kanji magical: each character carries a story, and when combined, they create layered meanings that English often needs whole sentences to express.

木 (ki) = tree → rooted, growing, living
林 (hayashi) = two trees → grove
森 (mori) = three trees → forest

One character: nature.
Two: community.
Three: abundance.

It’s poetry disguised as vocabulary.

Language as a Mirror of Culture

I used to believe that language simply describes culture.

Now I know: it shapes it.

The quietness between words in Japanese isn’t awkward, it’s intentional.
The honorifics aren’t formalities, they’re reflections of relationship, respect, and humility.
Even silence has grammar.

And that’s why learning Japanese doesn’t feel like memorizing verbs. It feels like learning how to be present.

When someone says 「お疲れ様です」(Otsukaresama desu) at the end of the day, they’re not just saying “good job.” They’re acknowledging effort. Sacrifice. Humanity.

No direct translation exists in English. Because no other culture puts quite so much weight on seeing someone’s struggle, and honoring it.

Why This Changed Me

Before Japanese, I saw language as a tool.

Now? I see it as a lens.

Learning kanji taught me patience.
Mastering particles like は and が taught me nuance.
Listening to native speakers taught me that sometimes, less is more.

And here’s the quiet revelation:

Japanese doesn’t ask you to speak louder. It asks you to listen deeper.

That shift, from trying to express myself to truly understanding others, has changed everything.

Not just how I communicate… but how I live.

A Gentle Invitation

If you’ve ever felt intimidated by Japanese, the characters, the grammar, the sheer depth of it, I get it. I was too.

But here’s what I’ve learned:
You don’t need to become fluent to begin appreciating its beauty.

Start with one kanji.
Write it slowly.
Ask yourself: What does this shape remember?

Try saying “arigatou” with intention, not just as a phrase, but as gratitude given.

And if you’re curious… go ahead.

Pick up a notebook. Learn 五 (go), five.
Then 九 (ku), nine.
Notice how they look nothing alike… yet both carry weight.

Language isn’t about perfection.

It’s about wonder.

Thank you for walking through this with me.

Arigatou gozaimasu, for reading, for wondering, for being here.

Sayonara for now…
but not goodbye.

🎌